REINFORCED ASPHALT OVERLAYS FOR PAVEMENTS - UNIVERSITY OF NOTTINGHAM DEPARTMENT OF CIVIL ENGINEERING doc

' Note that for the purposes of this document, the term 'reinforced asphalt' refers to asphalt layers that include grids and fabrics, and not fibre-reinforced materials... Chapter 2- Pr

UNIVERSITY OF NOTTINGHAM

DEPARTMENT OF CIVIL ENGINEERING

REINFORCED ASPHALT OVERLAYS

FOR PAVEMENTS

By

Paul John Sanders

Thesis submitted to the University of Nottingham for the degree of Doctor of

Philosophy October 2001

To my family

Thank you for the sacrifices you have made to give me the freedom to undertake this work

11

ABSTRACT

The maintenance of road pavements in England has become a costly necessity, due largely to the large volume of commercial vehicles using the roads which cause pavements to deteriorate quickly, and makes their repair more difficult to carry out These roadworks incur not only direct works costs, but also indirect costs from factors such as congestion, motor accidents and pollution There is obviously a need for cost-effective maintenance that minimises the occurrence and duration of these disruptions

To strengthen pavements bituminous overlays are often used, but may crack prematurely when placed over a layer with discontinuities such as cracks or joints, or deform excessively under wheel loading The problem of 'reflective cracking' is widespread and reduces the life of maintenance treatments considerably To increase the time before cracking appears on the surface of

a pavement, a (more expensive) thicker overlay may be used, but this can lead to problems with property thresholds and bridge clearance One possible option of reducing the thickness of overlays by making them more resistant against cracking and deformation, is to place a layer-of reinforcement within or

at the bottom of the overlay Although this approach has been used occasionally to reinforce overlays, over 40 years or so, it is not favoured with many road authorities, as the results of these treatments are difficult to anticipate, and may not be cost effective

This thesis describes an investigation into the effect of reinforcing thin bituminous overlays to identify key factors that significantly influence their performance By identifying these factors, optimum use of reinforced asphalt should be possible, and thus maintenance of the road network made more cost effective

The investigation was principally carried out in the laboratory using beam tests, shear box tests, tensile tests on reinforcement and large-scale wheel tracking tests 2-D Finite Element Analysis was used in the analysis of test results

Results show that properly constructed reinforced overlays can be between two or three times more resistant to cracking, and have less than half the permanent deformation of unreinforced materials

III

ACKNOWLEDGEMENTS

I would like to thank all those who have contributed to the research, whether it

be by providing assistance, encouragement, advice or funding

In particular, I wish to thank the following individuals and organisations:

Professor S F Brown and Dr N Thom, my supervisors, for their help and guidance, throughout the work

The laboratory staff at the university, especially Andy Leyko, Barry Broderick, Shane Malkin, Ehsan Sharegh and Mick Hutchings

The research consortium that included Netlon Ltd, ABG Ltd, 6-D Solutions,

Maccaferri Ltd, Scott Wilson Pavement Engineering and Bardon Roadstone (now Aggregate Industries UK Ltd)

The Engineering and Physical Sciences Research Council (EPSRC) for the majority of the funding

iv

DECLARATION

The laboratory work described in this thesis was conducted at the University

of Nottingham Department of Civil Engineering between December 1995 and April 2000 Subsequent to the laboratory work, much of the analysis has been carried out at my home in Berkshire

I declare that this work is my own and has not been submitted for a degree at another university

Beam Testing Pavement Test Facility

Numerical Modelling - Finite Element Analysis Economic Appraisal: Whole Life Costing

Guidelines for Design Summary and Overall Conclusions Proposals for Future Work

vi

1-11

Chapter 1- Introduction

1.1 General

For an economy to be successful and efficient, a freely flowing transportation

network is important In Britain, over the past half century or so, industry has increasingly used the road network to fulfil this function, with rail and waterways becoming less well-used However, as traffic levels continue to rise, the road network is becoming less able to fulfil the needs of the economy, which in turn leads

to (interalia) more expensive goods and services and environmental damage

To relieve the general problem of traffic congestion, the historical approach has been to build new roads However, this is now becoming less desirable, less economic and less environmentally acceptable in Britain and most other developed countries Unless traffic levels are reduced, therefore, it follows that road congestion will not be relieved and will probably increase as the road network increased travel times, traffic delays, increased pollution and enforced lower speed limits (for safety reasons)

From the above, it is appears that maintenance measures to arrest or delay road deterioration are required and should be quick to implement and long-lasting These help to reduce traffic congestion by both keeping the works period as short as possible and by increasing the period between maintenance treatments An added incentive for more effective (long-term) maintenance is the lengthening back-log of road maintenance as a result of a reduction in funding in recent years [1.1,1.2,1.3,

1.4,1.5,1.6] This reduction of maintenance budgets has led to some lengths of pavement requiring a structural treatment only receiving a superficial treatment, to ensure safety is not compromised The effect of postponing structural maintenance

in the short term is often an increase in the cost required to bring and maintain the road network to an acceptable standard in the longer term This problem has been recognised by the UK Highways Agency (HA) who are now committed to evaluation

of highway construction and maintenance in terms of Whole Life Costs [1.7], an approach that facilitates comparison of different construction and maintenance options The need for quick and effective maintenance treatments has also been emphasised with the advent of long-term Design, Build, Finance and Operate (DBFO) contracts, where efficient maintenance strategies can make the difference between success and failure

At present the most commonly used maintenance treatments include

f) partial or full reconstruction

In terms of their classification, a) and b) are not considered structural maintenance treatments, c) and d) may be considered as enhancing the pavement structure, but e) and f) increase the structural capacity of a pavement

Chapter 1- Introduction

Where a pavement structure requires strengthening, options to reduce overlay thicknesses are desirable for protection of the environment, and economy, i e to reduce the quarrying of aggregates, and to provide a pavement with adequate performance at reduced cost One approach to achieve this, that has been used to

a limited degree over the past 50 years or so, is asphalt `reinforcement" i e the inclusion of interlayer materials placed typically between an existing pavement and the overlay As described in Chapters 2 and 3 (the Desk Study), a range of materials are commercially available that reputedly reduce rutting and/or cracking These include grids, fabrics and composites (having elements of both grids and fabrics), which may comprise plastic, glass or steel As will be later seen, the option

of asphalt reinforcement is not a straightforward option, and the significantly different nature of some of these materials (produced to combat the 'same' defect or defects)

is indicative of problems in (i) characterising the nature and causes of cracking, and (ii) providing a solution to the problems

Historically, limited use of grids and fabrics has been made if compared to other approaches such as partial reconstruction or thicker overlays This is so for a number of reasons that have contributed to the general lack of confidence in their abilities It follows then that highway authorities are understandably reluctant to give approval to maintenance treatments that have a relatively short (if any) track record

in the UK This in turn makes it more difficult for performance data to be accumulated Accordingly, maintenance treatments using grids or fabrics are more often found on county roads than on trunk roads, whereas other more conventional solutions such as thicker overlays and bituminous mixtures incorporating modified binders, for instance, are usually adopted in preference on the trunk road and motorway network Another reason for the lack in confidence in using reinforced asphalt stems from reluctance to use it on the part of contractors, who, during construction, may encounter difficulties if they are not experienced in laying grids and fabrics

A brief investigation of the market relevance of the project shows that the current UK use of asphalt pavement interlayers (grids and fabrics) has an annual value of around 2.5 million pounds To be more meaningful, however, this figure needs to be considered with savings made due to reduced overlay thicknesses, or treatments resulting in fewer interventions in the future When considered in the light of the budget for structural maintenance of the Motorway and Trunk Road network of approximately 250 million pounds, it is understandable that it is still seen as a small market niche by manufacturers However, the market is likely to grow substantially

as the need for more cost-effective maintenance and alternatives to pavement reconstruction increases

' Note that for the purposes of this document, the term 'reinforced asphalt' refers to asphalt layers

that include grids and fabrics, and not fibre-reinforced materials

1-3

Chapter 1- Introduction

1.2.1 General

It follows that before effective pavement treatments can be devised and evaluated,

an understanding of the way in which pavements fail is required For brevity, the following discussion is restricted to the main modes of failure of bituminous surfaced roads, which include fully flexible pavements, flexible composite pavements and overlaid rigid pavements These structures are shown in Figure 1.1

Typically, pavements 'fail' in serviceability by developing poor riding quality (manifested by driver discomfort and measured by longitudinal or transverse unevenness), or becoming unsafe, particularly through reduced skidding resistance The deterioration of riding quality of a pavement is measured as unevenness of the surface which may be due to permanent deformation of bound materials or differential settlement of supporting layers An unsafe pavement on the other hand is normally one with poor skidding resistance This occurs when the pavement surface

is made smooth by the passage of traffic, an excess of bitumen in the surface (bleeding), or standing water Apart from an obvious design or construction fault with surface levels, ponding of surface water occurs in ruts, or is due to the settlement of the pavement support

Apart from loss of skid resistance the two most common symptoms of 'failure' of bituminous-surfaced roads are rutting and cracking, and these are discussed in more detail below As a general comment it is noted that 'failure' of pavements relates almost always to that of serviceability and not of 'destruction' as might be the case with other engineering structures

Cracking affects pavements detrimentally in various ways Initially, layer strength is lost which leads to overstressing of lower layers, consolidation and as a consequence, permanent deformation In addition to the reduction in strength (due

to less intact material), cracks provide access for water which softens unbound materials and reduces shear strength

Cracks and permanent deformation are normally attributed to traffic and/or environmental influences However, aspects of construction may also help induce problems such as when carriageways are widened For instance, cracking may occur at the junction of old and new constructions as a result of differential deflections across the vertical interface, which are due to differences in support of the old and new constructions The principal crack types associated with pavements are now described

1.2.2 Fatigue Cracking

Fatigue cracking occurs due to repeated applications of tensile strain which eventually overcome the resistance of the material This phenomenon may be considered as having two phases: (i) initiation and (ii) propagation Crack initiation can be considered as where the repeated application of tensile strains cause micro- cracks to join and form a macro-crack The continued application of tensile strains then causes growth and progression of this macro-crack through the material, which

is known as the propagation phase

1-4

Chapter 1- Introduction

then causes growth and progression of this macro-crack through the material, which

is known as the propagation phase

Crack initiation has been relatively well investigated and defined through functions that relate tensile strain to number of load repetitions (see References 1.8 and 1.9 for example) Crack propagation on the other hand is less understood and defined and is influenced by factors such as

" the type and amount of bitumen in the mixture,

" the amount and type of aggregate present,

" adhesion of the bituminous binder to the aggregate [1.10]

" the nature of applied load (traffic and environmental)

Some of these factors are difficult to quantify and so present problems in predicting performance

The three main modes of crack propagation which may be considered are shown in Figure 1.2, and are termed:

Mode I (opening mode)

Mode II (shearing mode) and

Mode III (a 'tearing' mode),

In practice Mode I cracking could be expected at the bottom interface of a bound pavement layer when loaded, and Mode II cracking might be expected in material bridging a crack or joint subject to differential movement Mode III cracking is perhaps more difficult to visualise but could possibly occur adjacent to wheel loads (longitudinally) in a pavement

In the classical pavement bending mode, crack initiation is normally expected to occur at the lower interface of a layer, although in thick bituminous layers (typically in excess of 250mm [1.11] as are found on many of Britain's trunk roads and motorways), cracking has often been found as a top-down phenomenon Also,

where pavements have rutted, longitudinal cracks may be found adjacent to the 'shoulders' These modes of pavement behaviour are illustrated in Figure 1.3

1.2.3 Reflection Cracking

Mechanisms of reflection cracking are complex and can be due to a combination of the movement of joints or discontinuities in the pavement beneath an overlay, and environmental influences at the pavement surface In general, the main contributor to vertical movement of cracks or joints is taken to be traffic loading Horizontal movements are assumed to be caused by differential thermal expansion and contraction of the pavement layers The magnitude of vertical movements depends on a range of factors including support to the layer being cracked and roughness of the crack faces (interlock) For thermal loading on the other hand, the severity of temperature gradients through the layers, plus the coefficient of thermal expansion/contraction of the different materials are of prime importance More detailed descriptions of reflective cracking are to be found in proceedings of the 4 RILEM conferences [1.12,1.13,1.14,1.15] and De Bondt [1.16]

Chapter 1- Introduction

Notwithstanding the considerable work carried out to characterise and solve reflective cracking problems, mechanisms still remain somewhat undefined For instance, reflective cracking occurring in Lane 2 but not on the adjacent Lanes (Lanes 3 and 1) or on the hard-shoulder (that was not trafficked) has been reported This type of occurrence highlights the complicated nature of predicting pavement performance in the field, even with 'normal' pavement constructions As described later, the addition of non-asphaltic interlayer materials within asphalt layers serves to give more difficulties in analysis and performance prediction

1.2.4 Rutting

Rutting occurs as a consequence of deformation of the visco-elastic bituminous materials and/or permanent deformation of materials supporting the bituminous surfacing An illustration of the stress-strain-time response of bituminous materials

is given in Figure 1.4 The permanent deformation of bituminous material is due to viscous flow of the bitumen, which in turn is a function of loading time, stress level and temperature With repeated loading, permanent strains accumulate and manifest as surface deformation, recognised typically through the appearance of raised shoulders Deformation of the whole pavement structure does not produce shoulders The two different mechanisms are illustrated in Figure 1.5

To help prevent 'excessive' rutting, analytical flexible pavement design generally uses relationships between traffic-induced vertical strains at the top of the subgrade, and the accumulation of permanent deformation to determine layer thicknesses [1.8 and 1.9] No deformation within the bituminous material is therefore explicitly taken into account The permanent deformation of materials supporting the bound layers

is due to excess stress being transmitted through the bound layers and is a deficiency of the overall pavement design

Failure of flexible composite pavements typically occurs when cracks in the supporting cement bound (CBM) roadbase are 'mirrored' on the surface or reflect through to the surface At this stage, the mode of failure is typified by quite regularly-spaced transverse cracks Cracking of the CBM occurs through initial shrinkage of the material (and is therefore dependant on the cement content used), and due to daily and seasonal changes in temperature The changes in temperature induce contraction and expansion in materials which largely depends on the type of aggregate used

With prolonged or heavy trafficking, flexible composite pavements may also develop

a network of irregular cracks as the CBM gradually breaks into smaller pieces which

Overlaid rigid pavements typically fail through reflective cracking, as slabs tend to move or 'rock' under traffic loads due to support under the ends of the slabs weakening Surface cracks are then opened up over joints which can allow the ingress of water If this occurs, further loading can force water and fine material out

of the cracks which is termed 'pumping' Pavement failure can then become very much quicker as support is progressively reduced and rocking is intensified, which opens cracks and gives access to more water, thus making the situation worse With reduced support, concrete slabs may crack irregularly, especially close to corners and edges, which can often be seen as diagonal cracks

1.3 Maintenance treatments

It follows therefore that any economic maintenance treatment should result in the number of load repetitions required to initiate and subsequently propagate cracks being increased, and the rate of permanent deformation reduced Where overlays are to be used there are several possible approaches to enhance their performance These include increasing the thickness of bituminous material, placing a Stress Absorbing Membrane Interlayer (SAMI), modifying the properties of the overlay (by addition of a polymer perhaps), or adding a reinforcing interlayer between the surfacing and the cracked/jointed layer

The options are now briefly described

An increase in the thickness of bituminous material has two main functions, namely, to reduce the strain on the lower interface and increase the time of crack propagation due to the increased distance from the point of crack initiation to the surface This has been well-used historically but can be expensive, and is not compatible with the increasing environmental concern Also, in situations where no further increase in levels can be allowed, this option is not feasible

A SAMI is a relatively soft layer placed between the old pavement and the new construction The function of a SAMI is to reduce stresses generated by movement

in the cracked pavement to a level that can be accommodated by the overlay A SAMI is typically 4-8mm thick and often comprises a blend of rubberised bitumen A summary of the findings of a study into the effectiveness of SAMIs has been given

by Mukhtar and Dempsey [1.17] which (inter alia) states that SAMIs have been more effective when used with flexible pavements than rigid pavements and that SAMIs perform better the thicker they are and lower stiffness they have Also, full-width pavement treatment has been more successful than local treatment directly over the joint/crack area

Modified binders in bituminous mixtures [1.18] have been used for more than a decade to improve resistance to rutting and cracking Probably the most common

1-7

Chapter 1- Introduction

way of bitumen modification is with the addition of Styrene Butadiene Styrene (SBS), Styrene Butadiene Rubber (SBR) or Ethylene Vinyl Acetate (EVA) polymers The addition of these polymers helps to enhance the performance of the bituminous mixture at high and low temperatures, which in general means that mixtures are less susceptible to rutting at high temperatures and cracking at low temperatures

Design rules that can reliably quantify the subsequent performance of pavements incorporating these materials are however not presently well-developed, and field trials [1.19,1.20] give a range of effectiveness Work aimed at quantifying the properties of polymer-modified bitumen and the subsequent effects on mixture performance is reported in Reference 1.21

The inclusion of grids, fabrics and composites within a pavement construction is another option that has been tried in various forms since the 1950s Although the first use of (steel) reinforced asphalt occurred in the 1950s, it is only in the past 20 years or so that a range of different glass and plastic materials have been used in any significant quantity Early work carried out at Nottingham [1.22] helped to confirm the potential of reinforcing pavements with grids Conventional engineering philosophy would normally suggest that there are two main reasons for placing interlayer materials within a pavement First of all, if the interlayer material is stiffer than the asphalt, it will reinforce the layer (if adequately bonded to the asphalt above and below) by carrying load that would otherwise be carried by the pavement Otherwise, if the interlayer stiffness is less or of a similar magnitude to asphalt, then,

to enhance the properties of the pavement it must provide stress-relief or similar to protect the pavement This may also include a `crack-stitching' quality where crack initiation is not prevented, but crack propagation is delayed

Note that for the remainder of the document, the term 'reinforced asphalt' refers to any sheet, grid or combination of the two, within layers of asphalt

Although it would appear relatively simple to define which mechanism applies and what contribution it makes to engineering performance, few reliable guidelines exist

In fact, basic questions regarding the type of material to be used as an interlayer, and where it should be placed in a pavement still remain It also follows that more detailed questions, such as the nature of the bonding (adhesion or interlock) and the effect it has on pavement performance and how to achieve it in the field are also poorly defined

The main focus of the work described here was to investigate the principal factors affecting the use and performance of reinforced asphalt Once these factors were defined, it was reasoned, this knowledge would be used to predict pavement performance and make it possible to choose appropriate treatments on the basis of sound engineering principles

As cracking has been a major contributor to pavement failure in the past, the project

is primarily directed towards reducing or solving the problems of cracking However,

as is later seen, the possible role for reinforcement in slowing the development of rutting is also addressed

It is recognised that the best method of establishing field performance is to monitor the behaviour of full-scale in-service pavements This of course is not feasible for

1-8

Chapter 1- Introduction

every experiment due to practical reasons such as the number of test variables which need to be investigated which would result in excessive cost Accelerated testing of full-scale pavements is another option but would also be expensive even if the apparatus was available Physical and mathematical modelling is on the other hand affordable and possible with the facilities at hand, and offers some advantages over large-scale testing

More discussion of the approach chosen to investigate reinforced pavement mechanisms is given in Chapter 4

A summary of the aims of the project is given:

" To investigate the use of reinforced asphalt in the UK in particular, and around

the world in general

mechanisms

" To model laboratory test results and to apply these models to field situations

" To summarise findings and produce guidelines for the selection and use of

reinforced asphalt

The thesis is structured as follows:

Chapters 2 and 3 comprise a general introduction to reinforced asphalt, how and where it has been used, and with what success Chapter 4 summarises the results

of the desk study and discusses possible approaches to the problem These range from monitoring in-situ applications of reinforced asphalt under real traffic loading, to laboratory testing of each of the components (asphalt, reinforcement and the bond between them)

Chapters 5 and 6 describe laboratory testing of the components of reinforced asphalt i e reinforcement and interlayer bond

In chapters 7 and 8 the development of test apparatuses used to test reinforced asphalt beams and half-scale reinforced pavements is described Chapter 7 gives details on beam tests, and Chapter 8 describes wheel tracking tests

The numerical modelling of reinforced beam and pavement structures using Finite Element Analysis (FEA) is described in Chapter 9 The potential benefits in terms of reducing crack propagation is illustrated

Economic appraisal of reinforced asphalt using the Whole Life Costing (WLC) approach is given in Chapter 10 In particular, the appraisal shows that careful consideration of the field situation is required before reinforced asphalt will be an economic solution to rutting or cracking

Chapter 1- Introduction

Chapter 11 provides guidelines for the use of reinforced asphalt derived from both work carried out in this project, and from results of full-scale trials described in the literature

Chapter 12 gives an overall summary and conclusions of the project, and Chapter

13 gives suggestions for future work to be carried out to investigate some of the

'unknowns' discovered during the present work

1-10

Chapter 1- Introduction

1.1 NEWS, New Civil Engineer, 30 November 1995

1.2 NEWS, New Civil Engineer 25 January 1996

1.3 CE NEWS, New Civil Engineer, 4 July 1996

1.4 Ty Byrd, New Civil Engineer 23 January 1997

1.5 Government Statistical Service (1999) Transport Statistics Bulletin, National

Road Maintenance Condition Survey: 1999

1.6 'News' World Highways, July/August 2000

1.7 Highways Agency Business Plan 1999/2000 St Christopher House, London,

1999

1.8 Brown, SF and Brunton, J M, (1985) An Introduction to the Analytical Design

of Bituminous Pavements (3rd Edition), Department of Civil Engineering, University of Nottingham

1.9 Shell International Petroleum Company Ltd (1978) Shell Pavement Design

Manual - Asphalt Pavements and Overlays for Road Traffic, Shell International Petroleum Company Ltd

, London, 1978

1.10 Read, JM (1996) Fatigue Cracking of Bituminous Mixtures Phd Thesis,

Department of Civil Engineering, University of Nottingham

1.11 Wu, Fenghe (1992) Assessment of Residual Life of Bituminous Layers for

the Design of Pavement Strengthening PhD thesis, Department of Civil Engineering and Building, The Polytechnic of Wales

1.12 RILEM, (1989) Reflective Cracking in Pavements: Assessment and Control

Proceedings of the RILEM Conference on Reflective Cracking in Pavements, Liege, Belgium, March 1989

1.13 RILEM, (1993) Reflective Cracking in Pavements: State of the Art and

Design Recommendations Proceedings of the Second RILEM Conference on Reflective Cracking in Pavements, Liege, Belgium, 1993

1.14 RILEM, (1996) Reflective Cracking in Pavements: Design and Performance

of Overlay Systems Proceedings of the Third International RILEM Conference on Reflective Cracking in Pavements, Maastricht, Holland, September 1996

1.15 RILEM, (2000) Reflective Cracking in Pavements: Research Into Practice

Proceedings of the 4th International RILEM Conference on, Ottawa, Ontario, Canada, April 2000

Chapter 1- Introduction

1.16 De Bondt, AH, (1999) Anti-Reflective Cracking Design of (Reinforced)

Asphaltic Overlays PhD Thesis, Department of Civil Engineering, Technical University of Delft, Holland

1.17 Mukhtar, MT and Dempsey, BJ (1996) Interlayer stress absorbing

composite (ISAC) for Mitigating Reflection Cracking in Asphalt Concrete Overlays Final Report, Project IHR-533, Illinois Cooperative Highway

research Program, Department of Civil Engineering, University of Illinois at Urbana-Champaign

1.18 Brown, S F, Rowlett, RD and Boucher, JL (1990) Asphalt Modification

Proceedings of a Conference on US SHRP Highway Research Programme: Sharing the benefits ICE London pp181-203

1.19 Butterworth, P and Whitely, DN (1996) Report on Inspection of Modified

Surfacings: M6Birmingham to Carlisle Motorway, County Boundary to Junction 37, NB and SB Carriageways Highways and Engineering Division, Civil Engineering Laboratory, Skirsgill Lane, Penrith, Cumbria

1.20 Nunn, M E and Potter, J F (1993) Assessment of Methods to Prevent

Reflection Cracking Proceedings of the Second RILEM Conference on Reflective Cracking in Pavements Reflective Cracking in Pavements: State of the Art and Design Recommendations Liege, Belgium, 1993 pp360-369

1.21 Brite/Euram II, (1994) Quality Analysis of Polymer Bitumens and Bitumen

Products by Microscopy Image Analysis with Fluorescent Light University of Nottingham, Nottingham, UK

1.22 Hughes, DAB, (1986) Polymer Grid Reinforcement of Asphalt Pavements

PhD Thesis, Department of Civil Engineering, University of Nottingham

FLEXIBLE, FLEXIBLE COMPOSITE &RIGID

(a) FATIGUE CRACKING AND CRITICAL STRAINS

FAILURE MODES AND CRITICAL STRAINS

FIGURE 1.3

DEVELOPMENT OF PERMANENT STRAIN

DUE TO REPEATED LOADING

Chapter 2- Problem definition - Desk Study

2-1

Chapter 2- Problem definition - Desk Study

(i) (ii)

a survey of organisations using or designing reinforcing products for reinforced asphalt in Britain, and

" whether the organisation had indeed used reinforced asphalt, and if so,

" what type of reinforcement was used

" whether problems were experienced during placement of the reinforced

asphalt

" performance rating of the reinforced asphalt layer

The organisations selected were thought to represent a cross-section of pavement- related organisations and included county and borough councils, consultants and contractors

The format of the questionnaire was intended to help define the use of the main types of pavement interlayers, i e fabrics, polymer grids, glass-reinforced grids and steel grids Also, definition of how frequently grids or fabrics are used, in what situation they are used, and whether they have been successful was sought

2-2

Chapter 2- Problem definition - Desk Study

The form of the auestionnaire used is aiven below:

Geotextile Polymer Grid Glass Steel Mesh (e g or composite Reinforced Grid (e g

Polyfelt) (e g NETLON or composite Roadmesh)

AR-1 or AR-G) (e g Glasgrid/

Chapter 2- Problem definition - Desk Study

The results of the questionnaire are summarised in Table 2.1 and Figure 2.1 Details of the type of reinforcement used were obtained from subsequent follow-up discussions

Table 2.1 Summary of auesti

Cate or g y Geotextile Polymer Gl ass St ee lM esh

grids ARIA AR-GA

B Good performance' was reported on rigid pavements (x2), new flexible

pavements (x1), and with overlays on weak flexible pavements (x1)

C 'Good performance' was reported on weak flexible pavements (x5),

and on new flexible pavements (x2) It appears that in some cases although problems were experienced during placement of ARI, once the material was installed, good performance was obtained

D 'Good performance' was reported on weak flexible pavements (x4), on

CBM base pavements (xl) and on overlaid rigid pavements (x1)

E 'Good performance' was reported on weak flexible pavements (x1),

and on overlaid CBM-base pavements (x1)

Of the 29 organisations that reported using, installing or designing reinforced asphalt, 23 were County and Borough Councils, two were consultants and four, Contractors

2-4

Chapter 2- Problem definition - Desk Study

The survey provided some interesting results, and trends in the performance

of particular types of reinforced asphalt became more obvious, especially during subsequent discussions with respondents to the survey However, a shortcoming of the investigation was that in many cases it was difficult to clearly define what was meant by 'clear benefit' or 'good performance' from either the survey sheets, or from subsequent follow-up discussions One of the main reasons for the lack of clear definition seemed to be, that unless problems are reported, engineers do not, as a rule, closely monitor sections that are not trial sections Indeed, it appears that in general, rating of the performance of reinforced asphalt sections tends to be subjective and little information additional to that supplied with the questionnaires was found to exist Also, in many cases, 'control' sections adjacent to the reinforced sections were not constructed, so objective comparisons of performance were difficult to make

Figure 2.2 portrays the data given in Figure 2.1 as percentages It is interesting to note that problems were only experienced on 4% of glass grid installations whereas the rate for steel grids is 28% However, this may be misleading as the 'Problems Experienced' category generally refers to problems experienced during installation Installation techniques have developed considerably in the past few years and have become more reliable, thus probably reducing the incidence of problems If the categories 'Good Performance' and 'Clear Benefit' are combined, polymer grids have the highest percentage However, the returns also showed that 'problems' were experienced with this category of reinforcement 21% of the time It is noted that the percentage of returns in the category 'Too early to tell' is significant, particularly for the glass grids Depending in which category these will eventually fall, the present distribution might be considerably different

A summary of the most significant points found in the survey is now given:

" The number of organisations (1 in 6) using steel reinforcement was

thought surprising as steel grids had not long been commercially available when the survey was carried out in 1996 This contrasts with polypropylene grids, for example, that had been used for a decade or more

" `Good performance' was reported with each type of reinforced asphalt, i e

polymer grids and fabrics, steel products and glass-reinforced products, in particular situations This is considered particularly interesting, as for this

to be the case, it is reasoned, different mechanisms of asphalt- reinforcement interaction must exist

" Polypropylene grids tend to perform well on `weak' foundations

" Polypropylene grids and composites and steel grids were prone to giving

problems during construction

2-5

Chapter 2- Problem definition - Desk Study

" Glass-reinforced materials were found to be quite popular and generally

performed well, but were apparently not as effective as the polypropylene materials on pavements with high deflections

Of the glass products, the proprietary product 'Glasgrid' was popular, partly due to its general performance, and partly due to the ease of installation (on account of the self-adhesive backing)

" In addition to the above observation regarding polypropylene grids on

weak foundations, there was a perception that polypropylene grids and composites were more effective than glass products where differential (vertical) movements at joints in concrete slabs were relatively large

" Geotextiles seemed to be effective over jointed concrete pavements It

was thought that this could be due to the bitumen-soaked geotextiles' ability to undergo high strains and reduce stress concentrations

" From discussions during the survey it appears that the relatively few

returns from consultants seems to indicate that the perceived risks (the absence of a recognised design method) in specifying reinforcement in asphalt are considered to outweigh the possible benefits Also, there seemed to be a general lack of awareness (or interest) among consultants concerning the use of fabrics or grids in pavements In addition, regional offices of the Highways Agency are often reluctant to take responsibility for accepting reinforced asphalt as a reliable option in maintaining the trunk road network

Assessment of the survey results and follow-up discussions with respondents is described as follows:

a) A considerable number of highway-related organisations have had

some contact with reinforcement in bituminous pavements

b) Each of the `main' groups of reinforcement is used in Britain

c) In general, limited knowledge exists of the performance of the

treatments used to date

d) Linked to c), measures of cost-effectiveness are almost always

subjective

e) Understanding of the way that reinforced asphalt 'works' is largely

anecdotal

f) Before reinforced asphalt is routinely used and accepted by

designers and highway authorities as a valid maintenance

2-6

Chapter 2- Problem definition - Desk Study

1)

treatment they must be shown to be effective, reliable and affordable

g) For reinforced asphalt to be accepted by highway authorities,

reliable design methods are required

h) To develop the design methods in g) good understanding of the

manner in which grids and fabrics influence pavement performance

is required

i) Results of well-documented full-scale (field) trials would help

provide confidence to organisations considering the use of grids and fabrics as maintenance treatments

The relatively high incidence of 'poor' performance appears linked

to problems during installation This is a practical issue that must

be resolved

As stated in f), for reinforced asphalt to be used routinely, the clients (highway authorities) need to be convinced that reinforced asphalt is an appropriate maintenance solution To provide this reassurance full-scale trials are preferred, but if this option is chosen, several years trafficking are normally required During this period, i e before the results of the trials are known, maintenance still has to be carried out, and so funding may be provided for unsuitable treatments Alternatives to quicken the production of results, and

so implement results sooner may be considered These include testing of small-scale samples in the laboratory or large-scale wheel tracking tests, which may be carried out in the field or in the laboratory Although it may be expedient to test small-scale samples in the laboratory, it is difficult to apply test results to full-scale situations Accelerated testing of test sections of pavements, on the other hand, provides test results that can be more easily applied to `live' pavements more directly and with more confidence Benefits

of using this approach include:

" several years trafficking can be applied in a few weeks or months,

" the magnitude of the load is known and can be varied

" the effects of temperature and moisture ingress can be accurately

monitored and (depending on the test configuration), controlled

Accordingly, for the reasons given above, accelerated testing of `pilot-scale' sections of reinforced asphalt pavement were carried out, and are described

in Chapter 8 Whilst not the preferred full-scale test, the Pavement Test Facility (PTF) was considered a worthy compromise, able to provide valuable insight into the performance of reinforced asphalt under wheel loading To supplement the relatively large-scale wheel tracking tests (by providing information on constituent materials), small scale laboratory tests were also carried out, as described in Chapters 5 to 7

2-7

Chapter 2- Problem definition - Desk Study

Appendix

Notes on Follow-up Discussions

Where possible, additional information to that supplied in the survey was obtained through telephone discussion Table 2.2 gives a summary of the main points of these contacts

Table 2.2 Summary notes from `follow-ups'

Reinforcement

Swansea Steel grid surface of the (possibly 80mm) overlay

Cracking and separation at interface with a Scottish Borders Poymer: Tensar 40mm overlay

AR1 No problems with 100mm overlay on grids

Can cause 'fatting-up' - probable migration Hampshire County Geotextiles of binder to surface

Council Polymer grids were found to be better than

Grids las rid in situations of 'high' deflection

Dorset County The grid 'moved' during installation of the

future

Applications have performed well on an

Polymer Grids area of haunch settlement and as a large

patch repair

Derbyshire Placing difficulties led to subsequent

debonding

Glass grid Also, under heavy quarry traffic over a

haunch, a 'thin' basecourse 'slid' on the rid

After initial problems with placing grids

Staffordshire Steel and (keeping materials flat), good performance

Polymer grids Wrekin construction Geotextiles All products were seen to slow-down but

Note 1 Problems were experienced with laying geotextiles, as they tended to 'pick-up'

on the wheels of construction traffic

Glasgrid was found to be the easiest to apply due to the self-adhesive backing All products present problems on roads with tight bends

2-8

Chapter 2- Problem definition - Desk Study

Table 2.2 Summary notes from `follow-ups' (continued)

Reinforcement

(Grid) the overlay' Suffolk County Poymer: AR-G Difficult to lay and overlay cracked Council (Composite) immediately

Worked well where a flexible pavement was Glasgrid widened with a rigid haunch Used to

remedy the problems noted above

Glasgrid Mixed success on flexible composite

Council Polymer grids Very effective on pavements with weak

subgrades See Note 2 below

Grids were placed on single-track roads

subjected to very heavy forestry vehicles Of

performed as well

Steel mesh is the other `referred' solution

Note 2: It was accepted that it was not possible to stop the pavements deflecting, but

the polymer grids hold the bituminous materials together

2-9

m

0

«

a

co (1)

0

0 F-

0

C,

C, (I)

ý

J

z

Chapter 3- Problem Definition - Literature Review

3-29

Chapter 3- Problem Definition - Literature Review

The main purpose of the literature review was therefore to further define the field performance of reinforced asphalt, and so help:

(i) establish the main influences on performance, and (ii) investigate the cost-effectiveness of reinforced asphalt

In addition to these findings, information on suitable test techniques and design approaches was also sought

Accordingly, information describing the behaviour of reinforced pavements is summarised below, and information relating to the investigation of laboratory performance of reinforced asphalt is provided in Chapters 5 to 8

In a similar way to the results of the questionnaire, the literature survey showed that there are a limited amount of well-documented 'complete' case histories that permit a full and fair comparison of the performance of different reinforced asphalt solutions to be made

The main reason for the lack of detailed case studies is typically that of cost,

as to obtain sufficient data from which to draw reliable conclusions, trial sections require expensive instrumentation and careful monitoring Also, considerable time is normally required to ensure sufficient trafficking has occurred for reliable comparisons with other reinforced, and unreinforced sections to be made In addition, where traffic is the prime cause of distress,

it is often difficult to determine the 'effective' traffic load that the section has experienced, bearing in mind factors such as traffic wander and variations in speed, axle load and dynamic effects

Some of the largest full-scale trials reported in the literature were carried out

in America These include; for example, the studies carried out in New Mexico [3.1], Pennsylvania [3.2], and Illinois [3.3] These studies summarise the assessment of around 13 different interlayer treatments on a large number (literally hundreds) of sites

Chapter 3- Problem Definition - Literature Review

Monitoring and evaluation of the trials indicated that reinforced asphalt could

be beneficial and cost-effective when incorporated into an overlaid pavement, although this was by no means true in all cases In one reference, [3.2] the authors actually comment that 'paving fabrics and fibrous treatments to retard reflective cracking are not recommended on the current analysis of life cycle costs' Similar studies in Europe [3.4 and 3.5] also show mixed results, with grids and fabrics performing both well and poorly in a variety of circumstances Accordingly, from this evidence the cost-effectiveness of reinforced asphalt is not obvious

It appears that the main reason for difficulties in determining the performance, and hence accurate appraisal of the cost-effectiveness of reinforced asphalt is the number of variables involved and their possible combinations The main variables include:

" pavement type (e g flexible, or composite)

" traffic volume

" traffic wheel loads

" the condition of the pavement before overlaying with reinforced asphalt

" pavement foundation support,

" differential movement of cracks and joints

" the type of fabric or grid used

" the position of the reinforcement in the pavement

" the type of bituminous mixture applied,

" the thickness of the bituminous mixture applied,

" the manner and degree by which the reinforcement is bonded to the

asphalt

" climate, (in particular annual and daily temperature changes)

Inevitably, each parameter will have a different influence on the overall performance of a reinforced asphalt pavement depending on the situation, so the number of possible combinations is large In assessing documented trials of reinforced asphalt, it is intended to identify the most important variables and thus enable a suitable investigation to be carried out

A summary of findings relating to the main types of reinforcing products is now given

Nunn and Potter [3.6] give the results of a road trial site built on an overlaid cracked concrete pavement to compare the performance of asphalt reinforced with grids with a control section After four years, the section built with a polypropylene Tensar grid was seen to have around 50% of the number of cracks reported on the control and HaTelit(polyester)-reinforced sections The principal cause of cracking was 'low' temperature (cracks were reported

to have initiated in the winter months)

Chapter 3- Problem Definition - Literature Review

Herbst et al [3.7] present the findings of trials on overlaid concrete pavements

in Austria where the performance of layers reinforced with geotextiles, geogrids, and a steel mesh, was compared with a control section and one treated with a SAMI (in this instance a layer of granulated asphalt beneath the binder course) The authors report that of the 60mm-thick asphalt sections trafficked, the Tensar-reinforced pavements (where a levelling course was used) behaved well, although the trial was inconclusive due to the limited (3 years) data The steel-reinforced sections cracked early in the trial, which was thought due primarily to installation (nailing) problems Early performance of the geotextile-reinforced section was good, although later in the trial, more cracking was noted

A simple economic appraisal was made comparing the cost of maintenance treatments, i e the difference between a 120mm overlay and 60mm overlays with the various treatments This showed that the geotextile solution to be cheapest, with the geogrid in second place followed by the conventional 120mm overlay However, performance was not explicitly taken into account

in the appraisal, which makes the comparison less useful The principal cause of cracking was not stated, but as the asphalt was placed on concrete, thermal effects were suspected

Gilchrist et al [3.8] describe the performance of a polypropylene grid-fabric composite (Tensar AR-G) used in a pavement on a soft foundation The composite was placed over a regulating layer with Hot Rolled Asphalt (HRA)

as a binder course After five years, the pavement was free of cracks, which compared well with the control section which cracked within four months The principal cause of cracking was that of relatively large movement of the foundation

Two examples from Germany [3.9], generally agree with results from the survey described in Section 2.1, i e that polymer grids seem to be effective in pavements with weak foundations

The first site was an approach road built over clay and peat in Schleswig- Holstein which was mainly trafficked by heavy vehicles Tensar grid was fixed

to the old pavement by nailing, and protected from construction traffic using a chip-and-spray treatment This was overlaid with an 80mm bituminous layer After 7 years of heavy trafficking no cracking was reported Previous to this, cracks reappeared in the pavement surface within two years of overlays being applied The principal cause of cracking was that of traffic-induced movements on a soft foundation

The second pavement described was built on a high embankment over peat and clay The road had been trafficked for 6 years to allow for the majority of settlement to occur before the reinforced overlay was placed, and was badly cracked The principal reason for using the reinforced overlay was to keep the thickness (and hence vertical load) to a minimum to reduce further settlement 80mm of bituminous overlay was then placed, and to improve the bond between the grid and the old pavement, a polymer-modified binder was used in a chip-and-spray process Monitoring over four years subsequent to

Chapter 3- Problem Definition - Literature Review

placement of the reinforced overlay showed the pavement to be uncracked

As for the previous case, the principal cause of cracking in the unreinforced pavement was that of traffic-induced movements on a soft foundation

The performance of a range of maintenance treatments applied to a flexible composite pavement was compared by Silfwerbrand and Carlsson [3.10] The treatments included polymer-modified binders (in the binder course), geogrids and a geotextile Results showed the section containing the geogrid to be the poorest performer, which was partially attributed to a poor placement technique The authors observed that treatments comprising 'homogenous asphalt layers' or of two layers of similar properties (stiffness) cracked least of all

Although the cause of cracking of the initial asphalt layer was not defined in the article, with flexible-composite pavements, a combination of shrinkage cracks in the roadbase and traffic loading is suspected

The performance of trial sections of the M6 motorway in England is described

by O'Farrel [3.11] As for the previous case history, the pavement treated had

a flexible composite structure, and the treatments -under investigation were combinations of fabric, geogrids and polymer-modified asphalt Inter alia, results of the trial showed that

" Rutting increased where a geotextile was present

" Placing the reinforcement between the roadbase and bindercourse was

more effective in inhibiting cracking than if the reinforcement was placed between the binder course and the surfacing

" GlasGrid was more effective laid in 'larger' continuous areas than strip

treatments

" The GlasGrid was easier to place than the polypropylene grid

As for the previous case history, the cause of cracking in the asphalt surfacing

is thought to be a combination of traffic loading and shrinkage cracks in the CBM roadbase

A road trial comprising 17 sections in Sweden, reported by Johansson and Ancker [3.5], compared the performance of geogrids, geofabrics and control sections on a thin pavement (350mm from top of subgrade to the surface) The construction was that of a grouted macadam base under a 50mm surfacing, which had suffered extensive longitudinal and alligator cracking The geogrid suffered from problems with installation, which led to considerable cracking early in the trial Other treatments, on the other hand were monitored for six years and showed (inter alia) that the polyester geotextile performed considerably better than the polypropylene fabric This was attributed to the larger amount of binder used with the polyester, which suggests that the binder-soaked geotextile functioned as a stress-reducing layer, similar to a SAMI This might suggest that the main function of the geotextile is a bitumen reservoir

The alligator cracking in the 'original' pavement was attributed to binder aging, but no reason for the longitudinal cracking was given

Chapter 3- Problem Definition - Literature Review

Yaromko et al [3.12] report the findings of maintenance treatments applied to roads in Belarus that employed geotextiles, grids and control sections It appears that the geotextile-reinforced sections fared better than the polymer grid-reinforced sections, although all grid and geotextile treatments were found to reduce cracking

In summary, although little detail is given in most of the references, there is a suggestion that polypropylene grids are more effective in crack reduction when used in a situation where vertical movements are 'high', such on a soft foundation This contrasts with situations on rigid or flexible composite pavements where movements are often expected to be thermally-induced, and thus tending to be horizontal However, although often performing well once installed, polypropylene grids were generally found to be more difficult to place successfully than some other reinforcement types

Five case histories giving examples of the application of glass-reinforced composites and grids to fully flexible, flexible composite, and granular base pavements have been described by Doligez and Coppens [3.13] The composites were positioned below the wearing course, i e covered with between 20 and 80mm of asphalt, and to gauge the effect of the maintenance treatments, Benkleman Beam and LaCroix Deflectograph tests were carried out before and after maintenance was carried out Rut measurements on some of the sites were also quoted

Test measurements taken before and after application of the composites showed both rutting and transient deflections to be less than the unreinforced pavements In addition, deflections actually reduced on two of the pavements

as time passed, suggesting that the condition of the foundation had improved The authors quote an engineer responsible for the road as saying that 'it seems that the stiffness of the glass fibre distributes the stress and helps the soil structure to become stable', but no detailed suggestion was given as to how this actually occurred

From the article it appears that the glass-reinforced interlayer is effective on relatively thin roads subjected to heavily loaded axles However, as a cautionary note, the authors make the point that vertical shear must be considered before specifying glass-reinforced grids Although not specifically stating why, it is assumed that these relatively brittle grids are susceptible to this type of failure Also, (and common to many of the references encountered during the literature review) correct placement of the grids was stated as being very important to obtain good performance

Reference 3.12 reported that glass-reinforced grids generally performed well

on Roads in Belarus However, it is not obvious on what road structures the materials were placed, or what the mode of cracking was

As mentioned in Section 3.2, trials on the M6 motorway in England [3.11] showed that GlasGrid applied in 'continuous' layers was more effective than

3-6

Chapter 3- Problem Definition - Literature Review

GlasGrid applied in strips Also, the application of this adhesive-backed product was found to be more straightforward than the use of polypropylene grids

Examples where geotextiles have been used successfully can be found on most pavement types, and they are more widely used than any other type of interlayer product This is largely due to their relative ease of application, their (perceived) generally 'good' performance, and 'low' price per square metre

A key difference between fabrics and grids is the absorption properties of the fabrics for soaking-up bitumen This results in a tendency to act like a SAMI (stress absorbing membrane interlayer), absorbing stresses, rather than adding strength to the pavement This being the case, it is understandable that geotextiles have been reported as not performing well under load on soft foundations However, as is seen below, this is not always the case, as geotextiles have been successfully used as part of thin surfacings on 'low volume' roads in Australia, although for this usage, the primary function is more to waterproof the pavement, than to act as a SAMI

A selection of references describing the use of geotextiles is summarised and given below

A study in New Mexico [3.1] compared 7 different interlayer treaments on six sites on a mixture of fully flexible and flexible composite pavements The treatments investigated were non-woven fabrics and rubberized asphalt, and were compared to the performance of control sections Results showed that in general, interlayers can retard the rate of reflective cracking and therefore make savings on maintenance costs However, the benefits of the treatments were not clear-cut Mention was made of the importance of good construction control, as problems were experienced in laying some of the products, largely because of insufficient preparation of the pavement surfaces and incorrect amounts of 'bonding coat' (binder) being applied

A field experiment in Pennsylvania [3.2] was set up to compare the relative performance of four different fabric interlayers, a fibre-asphalt interlayer, a polyester fibre-reinforced asphalt overlay and a control section The pavement had a mixture of fully flexible and overlaid concrete construction Problems were encountered during placement of materials, due mainly to 'contractor inexperience' but also to the nature of a heatbonded fabric which tended to 'wrinkle' and move under site traffic

Significantly, after 44 months none of the treatments were considered cost- effective, (compared on the basis of surface cracking), but it is fair to say that this area is prone to low temperatures and wide thermal cracks against which geotextiles are (reputedly) not effective

In the UK, Walsh [3.14] reported good performance for geotextiles over cracked concrete slabs, especially when used with an HRA overlay A

Chapter 3- Problem Definition - Literature Review

potentially important point is made concerning the need to limit the amount of vertical movement of the slabs before application of the geotextile In this case, it was achieved by grouting voids beneath the slab

A comparison of the costs of geotextile (£1.30/m2) and an 'equivalent' 40mm overlay (£4.00/m2) was given However, it was also reported that until both the untreated and the geotextile-treated sections had cracked, a fair comparison could not be made

Other instances where geotextiles have been used over concrete or flexible- composite pavements are given in references 3.15 to 3.21, and in general, the results of the trials indicate good performance However, there are some exceptions, such as Reference 3.20, where more rutting was recorded in the fabric-reinforced section, and in Reference 3.19 where cracks were reported

as reflecting through the fabric-reinforced layer relatively quickly This was attributed to the concrete (over which the reinforced asphalt was placed) not being cracked and seated prior to laying the reinforced overlay

Apart from the stress-reducing function of a geotextile, the waterproofing properties may be very important, as the papers by Dumont and Decoene [3.16], Van Deuren and Esnouf [3.22] note A paper by Phillips [3.23] also describes the use of bitumen-impregnated geotextiles to seal pavements with thin structures on expansive soil subgrades These papers serve to highlight another positive feature of fabric behaviour - their ability to undergo large strains without 'failing' Also, if cracks do reflect through the wearing course, the bitumen-soaked material still remains watertight Use of geotextiles to stabilise moisture contents of low volume roads could also be extended to higher grade roads where water-susceptible materials are present

Barksdale [3.24] collated and summarised a large amount of information from field trials in the USA and concluded the following:

In general fabrics were found to be more effective in reducing the incidence and severity of reflection cracking in temperate or warm climates In some cases the waterproofing qualities of bitumen-soaked fabrics were noted as being the most important benefit of applying fabrics

For flexible pavements:

Moderate to significant levels of reflection cracking could be delayed by

2 to 4 years by using a full-width fabric interlayer

" Fabrics were found to be most effective where tight closely-spaced

cracks were found (similar to `alligator' crack patterns)

The limit of crack width for successful application was 10mm

Fabrics were not effective where thermal cracking was a problem (as these cracks may often be greater than 12mm wide)

Mukhtar and Dempsey [3.25] note that geotextiles tend to perform better on flexible pavements that exhibit distress via closely-spaced alligator (fatigue) cracking, rather than on pavements with large cracks and/or large deflections Limiting values for use of geotextiles are given in Table 3.1